Is Political Analysis a ‘Scientific’ Enterprise? Should It Be?

According to Fay (1996)¹, the study of political science involves a close analysis of dates, governments and politics. This discipline concerns with the study of human behavior and all issues related to politics as observed in controlled environments. This means that it would be very difficult to reproduce results through experimental methods.

Many scholars have argued that the study should be based on observations rather than experiments. Historically, most advances and conclusions in political science are derived from observations of the general trends of nation’s politics and government.

The fact that the results from studying a variety of governments of a given number of nations limits the probability of obtaining reproducible results and thus limits the study of political science to be classified as a scientific enterprise. Political analysts utilize a variety of means to obtain their research data. Most primary data is obtained through historic records and official documents.

The secondary materials have been traditionally derived from results of surveys, case studies, statistical analyses and experimental models and research buildings. Political science builds its theories on critical observations of political elites, individuals and groups, as well as institutions. These patterns are drawn, generalized before the theories are being created.

The correlation between the study of political science and science is a subject to debate due to some specifications that scientific studies and models stipulate as mandatory in all scientific models. This paper analyzes some of the scientific requirements in a scientific enterprise and then correlates them to the model and research methods used by political analysts to draw its conclusions

Literature review

The definition of political science has changed over the years since the 1950s. Initially, the discipline had been comprehensively descriptive in nature, but has since relied on statistical methods. Political analysts have over the years rejected political science as a science based on two approaches: practical and philosophical objections.

In the first one believes that political behaviors are very complex and therefore mislead researchers having made the data collection quite hard. Philosophical objections, on the other hand, base their arguments on the fact that human reasoning cannot be objectively quantified. This leaves observations and perceptions by the researchers as the only alternative (Rosenberg, 2012)².

Political analysis lacks a standard methodology for quantifying the validity of its measurements and therefore makes it hard for obtaining testable results. The complexity and uniqueness of the political analysis make the results of the study unreliable. According to Cherry (2010)³, reliability is a measure of consistency of a given measurement. For a test to be reliable, it must deliver similar results on a regular basis.

The results are independent of the reason for administering the test, but should be approximately similar every time the test is done (Cherry, 2010)⁴. Unfortunately, mathematical quantification is not possible in political analysis when making approximations about the reliability of a given test.

Two aspects are normally considered in science: inter-later and internal reliability. In order to assess the first one, a minimum two independent judges should score the test. The scores are then analyzed to establish the consistency of the rater’s approximations. “The best way to test inter-rater reliability is to assign each rater to test a score” (Silva, 2010)⁵.

Then, the test administrator computes the correlation between two ratings to establish the degree of inter-rater reliability. On the other hand, internal reliability is a measure of consistency of test results conducted on the same test. The researches compare test variables, which measure the same construct to establish the tests internal consistency(Cherry, 2010)⁶.

On many occasions, these tests involve very similar or repetitive questions that are utilized to assess the reliability. Since the questions are identical and are structured to answer the same phenomena, the researcher must provide answers to the question in a similar fashion meaning that the test has internal consistency (Cherry, 2010)⁷.

Political analysis lacks a standard methodology for quantifying the validity of its measurements and therefore makes it hard to obtain testable results. Additionally, science capitalizes on knowledge to explain how phenomena is the way it is. As a result, theories and testable predictions based on empirical reliability are created.

The more similar observations are observed, the more the theory is deemed to be testable and therefore correct. Poor consistency of the human behavior thus reduces the rate of testability of the theories and consequently reduces the reliability of the results (Silva, 2010)⁸.

Political analysis is a social study and involves numerous examples of complex human behaviors. Human beings are neither identical nor interchangeable. Their complex nature makes them lack a general conclusion based on the results of the methodology used.

In science, generalization is a must concept and relies on the building relationships between two or more facts and utilizes empirical generalization to reach a credible conclusion (Buttolph & Reynolds, 2012)⁹.

Since humans cannot be put in a Petri dish in an effort to control their behaviors to study certain variables, the validity of the results obtained is thus questionable. It would be important to look at the definition of variability as an important factor in quantifying scientific studies (Silva, 2010)¹⁰.

Validity is defined as the ability of an instrument to achieve its goals. Tests should be valid in order to ease interpretation, and provide accurate application of the results. Validity is a measure that uses research methods that can demonstrate the existence of a predefined relationship between the test and the behavior it is supposed to measure. There are two types of validity measures: content and face validity.

In case the elements of the test represent the whole range of possible results the test should cover, then the test have content validity (Silva, 2010)¹¹. Content validity draws the test questions from a wide pool of items covering a wide range of topics.

Face validity, on the other hand, is a component of content validity that is determined when the person assessing the instrument makes a conclusion that measures the desired element or trait in the test (Cherry, 2010).

Quantification

While scientists insist on quantifiable tests, Political analysis, being a social science, relies on behaviors, values, beliefs and so on to make conclusions. This implies that their results cannot be measured accurately. Measurement is a fundamental element of science and although generalizations and assumptions are common, the quantification process must be illustrated clearly.

Some political scientists vehemently criticize this approach by citing the scientific methodologies that have been employed historically in science. Galileo, for instance, used a crude telescope to view the moon and other planets. Scientific developments in astronomy were developed from the foundations he discovered through more and more precise observation instruments.

Another fundamental element of science is the ability to offer transmissible knowledge. Scientists insist that scientific knowledge must be transmissible or, in other words, they must be explicit enough to allow other scholars to analyze them and replicate their results. Unfortunately, the methodologies used in social sciences are basically based on objective and systematic observations.

This means that a fundamental goal of science is left out since the results obtained cannot be verified. The theories constructed from these verifications of public inspection and results are not in a credible position to explain why phenomena behave the way it does (Fay, 1996).

In the scientific world, the knowledge is used to describe things and facts, for example, how things are created and why they are what they are. On the other hand, political analysis depends on causal relationships which derive their results from correlations in order to establish useful and important information. These however fail to explain issues related to why and how (Fay, 1996)¹².

Falsifiability

During the development of theories, it must be quite clear whether the theory is falsifiable or not. Science insists that an empirical theory fails to satisfy the fundamental of science if it cannot be proved to be either right or wrong through various tests. When a theory is found to be wrong, it gives the researchers the motivation to look for alternative theories through elimination of the previous theory.

The truth is that no political setting is interchangeable. If we argue from this point, it means therefore that the results we will obtain from analysis of a political setting will be subjected to alterations once an identical study is carried out.

This would consequently lead to situation where theories will be formulated today and dismissed the following day. Political analysts use case studies to come up with comprehensive theories but fail to generalize them being afraid of criticism from independent similar theories.

‘The Organization of the American Political Science Association’ (2006)¹³ reported new developments in the scientification of political analysis. In the early 1960s, most political analysts used scientific means to build on the behaviorism during the early foundations of political science as a discipline.

However, over the years, these theories have been challenged by different authors owing to their nature of lack of reliability and generalization. The line between facts and values has been a nightmare to most political scientists and other social scientists for a long time. Some held on the argument that values are facts, but in the real scientific world, a clear distinction between the two must be established.

Epistemology

As one of the key pillars of philosophy, epistemology concentrates on propositional knowledge. The political analysts utilize this approach by classifying certain elements of beliefs and values to be true or false. The results produce normative knowledge which deals with the evaluation and recommendation of what should be rather than what it is, or how it came to be.

On the other hand, real science aims at creating a non-normative kind of knowledge which is based on facts or, in other words, the objective determination of what things are. While the real scientific world insists on knowledge being cumulative, the research methods used by previous experiments and results should be used as foundations of new work.

Political analysts deal with variation in the settings of political arenas, governments and institutions in such a way that the conclusions from different studies yield different results. In their words, the repeatability, reliability and validity of the data obtained cannot be used to make a general observation.

By presenting a systematic a systematic and well reasoned anticipation, science is able to make accurate predictions for future occurrence of events. Thus, science is in a position to generate predictions for the future. Social analysis fails in generalization of its theories and therefore most of its conclusions cannot be used to predict the occurrence of future events (Fay, 1996)¹⁴.

Science relies on parsimony in an effort to give other researchers a chance to choose between alternatives. Scientists therefore prefer methodologies that give a vivid explanation of phenomena while maintaining few parameters. In real scientific research, deductive arguments must not necessarily be reached.

These relate to arguments that are proved to be true “through the use of a series of logical statements to derive a conclusion based on true premises” (Buttolph & Reynolds, 2012)¹⁵. Although deductive reasoning is used in political analysis, inductive reasoning prevails in most research work making inferences though observations and propositions (Buttolph & Reynolds, 2012)¹⁶.

This brings the issue of probabilistic explanation. In science, it is required that 100% accuracy be met prior to making any predictions. The use of inductive reasoning therefore forbids this practice of making prediction based on previous findings. In other words, every situation in political analysis requires an independent research and conclusion.

For example, if a researcher was to research about the political situation in a country, say Britain, it would be wrong to make inferences about the United States using the results of the study. On the other hand, a scientific experiment could be reproduced to another similar case.

For example, the behavior of bacteria inside a Petri dish could be used to make general conclusions about the behavior of all similar bacteria. From the given example, several observations can be made. The situation of political institutions in Britain and America are likely to change with time.

By the time the study was taken, several parameters such as the nature of economy and the general perceptions of the respondents will have changed. The scientific study of bacteria in the same environment and conditions will yield similar results day-in-day-out.

We therefore concluded that the Petri dish experimental results are reproducible, valid and reliable. Reliability is important in this case in order to assess the practicability and efficiency of the materials used to gather the data (Martin, 1994)¹⁷.

Conclusion

The paper analyzed the scientific approach of research and made equal comparison to the study of political science. The scientific approach look more tangible and reflective results compared to the social science approach. The paper concluded that political analysis therefore cannot be classified as a scientific enterprise. For a long time, the study has based its reflections on descriptions obtained through observations.

The fact that these systems are undergoing transformations to include statistical inferences suggests the birth of a new approach to the subject.

At the moment, the scientific requirements for a theory to satisfy the discussed expectations of a scientific study do not allow the prevalent method in political analysis to be described as a scientific approach. Deductive reasoning may not be practical in political analysis since every time the research is carried out, the probability of obtaining different data is high.

Bibliography

1. Buttolph, J. & Reynolds, T. H., 2012, Political Science Research Methods, 7th edn., sage Publications. Web.
2. Cherry, K., 2010, Reliability and Validity. Web.
3. Fay, B., 1996, Contemporary philosophy of social science: a multicultural approach. Blackwell, Oxford.
4. Martin, M., 1994, Readings in the philosophy of social science, MIT Press, Cambridge.
5. Rosenberg, A., 2012, Philosophy of social science, Westview Press, Boulder, CO.
6. Silva, E., 2010, Racism without racists: color-blind racism and the persistence of racial inequality in the United States, Rowman & Littlefield Publishers, Lanham.
7. ‘The Organization of the American Political Science Association’, 2006, Proceedings of the American Political Science Association, Vol. 1, pp. 5-15, First Annual Meeting, 1904.

Footnotes

¹ Fay, B., 1996, Contemporary philosophy of social science: a multicultural approach. Blackwell, Oxford.

² Rosenberg, A., 2012, Philosophy of social science, Westview Press, Boulder, CO.

³ Cherry, K., 2010, Reliability and Validity.

⁴ Cherry, K., 2010, Reliability and Validity.

⁵ Silva, E., 2010, Racism without racists: color-blind racism and the persistence of racial inequality in the United States, Rowman & Littlefield Publishers, Lanham.

⁶ Cherry, K., 2010, Reliability and Validity.

⁷ Cherry, K., 2010, Reliability and Validity.

⁸ Silva, E., 2010, Racism without racists: color-blind racism and the persistence of racial inequality in the United States, Rowman & Littlefield Publishers, Lanham.

⁹ Buttolph, J. & Reynolds, T. H., 2012, Political Science Research Methods, 7th edn., sage Publications.

¹⁰ Silva, E., 2010, Racism without racists: color-blind racism and the persistence of racial inequality in the United States, Rowman & Littlefield Publishers, Lanham.

¹¹ Silva, E., 2010, Racism without racists: color-blind racism and the persistence of racial inequality in the United States, Rowman & Littlefield Publishers, Lanham.

¹² Fay, B., 1996, Contemporary philosophy of social science: a multicultural approach. Blackwell, Oxford.

¹³ ‘The Organization of the American Political Science Association’, 2006, Proceedings of the American Political Science Association, Vol. 1, pp. 5-15, First Annual Meeting, 1904.

¹⁴ Fay, B., 1996, Contemporary philosophy of social science: a multicultural approach. Blackwell, Oxford.

¹⁵ Buttolph, J. & Reynolds, T. H., 2012, Political Science Research Methods, 7th edn., sage Publications.

¹⁶ Buttolph, J. & Reynolds, T. H., 2012, Political Science Research Methods, 7th edn., Sage Publications.

¹⁷ Martin, M., 1994, Readings in the philosophy of social science, MIT Press, Cambridge.